Today, we already have the ability to manipulate the genetic code of various organisms. In the very near future, we will have the ability to directly alter the DNA in the cells of our bodies in order to cure genetic diseases. The ethics of gene therapy for consenting adults is not in question. This kind of gene therapy is just like any other form medication. As long as the potential gene therapies go through proper testing, it's fine.
The ethical questions arise when we talk about altering the DNA of germ cells, which are the cells that are involved in reproduction. Is it moral to alter the DNA of unborn children?
The answer depends entirely on how well we understand what the consequences of genetic manipulation will be.
If we are altering the DNA in order to correct a well-understood genetic disease, such as cystic fibrosis, then it is not immoral to correct the DNA. We know what the result will be, and it is a good result. A child will be born without a life-threatening genetic disease. Any genetic disease that is the result of a single base-pair mutation or is the result of one specific allele of one specific gene is relatively simple to understand. Thus, these genetic diseases can be safely corrected with genetic engineering.
If we are altering the DNA in order to accentuate a perceived positive trait such as intelligence or beauty, chances are that we don't know what the hell we're doing. Traits such as intelligence are determined by a multiplicity of genes interacting at levels that go beyond the base genetic code. Epigenetics is another layer of coding on top of the standard DNA base-pair coding that is still not well understood. If we lack the computational power to determine exactly what the result of the genetic alteration will be, then it could be that the baby will be born with a heightened risk of cancer, or will develop some sort of debilitating disease down the road. In other words, if we don't know what we're doing, instead of producing a positive trait, we could end up with a man-made genetic disease. If this baby survives into adulthood, there is a chance he or she will reproduce and introduce this new genetic disease into the general population. This is not something we want to toy with. It is immoral.
If it is immoral to experiment with poorly understood gene therapies with the intention of producing a human baby, how would scientists ever learn to develop gene therapies that might be beneficial? It's catch 22. You can't genetically engineer a baby until you understand what you're doing, but you can't understand what you're doing until you genetically engineer a baby.
With pharmaceuticals, researchers can test new drugs on animals before moving onto human trials. Even when the drug gets to human trials, that is usually done with the consent of the human subjects. Unborn children cannot provide consent. Moreover, from a practical standpoint, it is impractical to wait for them to grow into adulthood in order to measure what the impact of the genetic engineering will be. The researcher would be long dead of old age.
But there is a solution. If computational power becomes sufficiently advanced, it will be possible to predict the result of a genetic modification. If the computational models accurately and consistently predict the results in animal studies, then the computational models could be applied to humans. Even then, the researchers need to be willing to accept the consequences if their models turn out to be wrong. If their meddling produces a debilitating man-made genetic disease in the newborn child, then they are responsible for making sure the child is well taken care of from the cradle to the grave. If they refuse to take responsibility, then the punishment should be that the researchers will be exposed to radiation that alters their DNA. Remember: what you do unto others, others may do unto you.